[OpenLayers-Dev] Entering OSGeo Project Sponsorship

[Christopher Schmidt]

On Fri, Oct 17, 2008 at 05:48:50PM -0400, Erik Uzureau wrote:
> So first of all, the link to the Project Sponsorship page on the OSGeo
> wiki is great -- I probably should have looked at that more before
> writing anything. Alas...  :-)
> 
> But so one thing you mention in here is the GDL/OGR experience. Is
> there some documenation about that somewhere or is it more just
> firsthand knowledge. I think that knowing more about how this
> sponsorship deal has actually worked out for other projects would be a
> huge help to informing us as we put our feet into it. Any informal
> observations or links to more official documentation would be
> appreciated.

So far, I don't think there's anything like that, though I see Paul has
emailed GDAL/OGR requesting any info they're willing to share. Most of
my experience has been by asking on IRC and paying attention to their
project.

> Overall, it seems like the main tension between our points of vista
> here are centered around the notion that sponsors would somehow be
> hijacking the direction/focus of development activities for the
> project.
> 
> Although I think you know that I would resolutely agree with you about
> this in principle, it's not something I'm nearly as ardent about
> defending... so it's great that you wrote this, that there's someone
> out there clearly voicing this important issue.
> 
> Whereas I definitely agree, again, "in principle", I think that this
> is maybe some grey territoire here and that before making a public
> move with this, it would be good to sort it out.
> 
> The deal being that when I try to put myself in the shoes of Joe the
> Plumber ^H^H^H^H^H^H^H Sponsor, I unfortunately just don't think they
> will universally understand this vibe. In an ideal world, they all
> would and we could merrily go on about our business, but I just think
> that the typical "sponsor" will be more of the mindset "I give you
> money... what do I get out of it".  And even if we clearly explain to
> them the idea is just a general warm feeling of "supporting the
> project" combined with the logo/marketing buzz.... I think they'll
> still at least *feel* like maybe they have some sort of implicit vote
> or weight in terms of project direction.

I think that one thing to consider here is perhaps not all sponsors are
the kind of sponsors we're looking for at this time. Anyone who feels a
need to have direct involvement with project direction to be a sponsor
is probably not neccesarily who we want to target as a 'first pass'
sponsorship target: If we find that there are a lot of these as time
goes on, we can rethink this, but I think there's a number of smaller
targets we can take on which *won't* feel the need to get this kind of
invovlement, and would be happy just supporting the project. I'd rather
target these first.

> ...and now that I've written all that out, I'm not even really sure
> exactly whether I have any sort of a solution to that or not... just
> that it seems like something we should be attuned to and that if there
> were some way to somehow mitigate this factor, it might be worth
> looking into. if others have opinions/experience, please chime in.
> 
> One thing that I saw on that wiki for the Project Sponsorship thing
> chez OSGeo is the "Earmarked Sponsorships" [1]. Maybe there is someway
> that this could be a sort of intermediary happy ground in between
> these extremes. What I would really like is something like this but on
> the other side... like instead of PSC suggesting special projects
> looking for money, it would be companies suggesting special projects
> and offerring money.
> 
> I realize that your answer to this is "if a company wants something
> done, they should pay someone to do it" but I think we can look at the
> last 2.5 years and clearly see that that is *not* happenning. The only
> way that patches are getting into trunk is when a PSC member is
> interested in the patch (or someone out there makes a whole whole
> whole lot of noise about it).

I don't think that's entirely true. The vector behavior work is, I
believe, funded in some part by customers behind organizations working
on it, and things like the PointTrack layer were (I believe) originally
developed as customer requests and integrated into trunk. I think
there's enough evidence that this is possible that we can stick with
that assumption unless we see more evidence otherwise. 

> So again, I'm not saying "let's sell our souls to the highest bidder",
> all I'm saying is that maybe it would make sense to help out companies
> that find themselves in this situation but don't seem to be able to
> connect the dots to make this happen. Remembering, afterall, that the
> whole idea here is getting ideas and work out into the open so that we
> don't have 5 people doing the same thing (and especially if they  are
> all *paying* to do it :-P)

I think that we can work to improve the ways in which we explain to
companies who they can pay -- by helping to improve the OSGeo Provider
search, for example, or other similar mechanisms. I don't think project
sponsorship is neccesarily the right path for this type of problem.

> Another issue you bring up is the ideas for how to *spend* the
> [theoretical] money. I suggest prioritizing it for bunkers, whereas I
> *think* you are saying it would be better spent by funding people to
> do more administrative-type work for the project like bug tracking or
> email list question fielding or something of that nature. ("better" is
> maybe not what you're saying but at least the idea that it is
> something to be considered equally with the bunkers).
> 
> And again, I think I agree with this in theory, I'm just curious as to
> how the actual implementation of this would work out. I guess it all
> really boils down to how much theoretical money we are really talking
> about. If we were to somehow get like 10 gold sponsors for the
> project, then ya ABSOLUTELY it is time to start funding someone to be
> the bug man or the users@ list man (or woman, sorry). But if we're
> talking about $3,000 or somewhere more in that neighborhood, we're
> talking about something like $50-100 a week there... which I think is
> hardly enough to pay anyone to be actually productive in an
> administrative/maintenance role. Let alone the idea that anyone would
> even be *interested* in it, jeje. :-)

Sure. The idea of sponsorship funding a maintainer is definitely based
on the idea that we have a handful of silver sponsors: not enough that a
full time salary like yours or mine could be paid out of it, but perhaps
someone more junior could do it as a part time job. GDAL/OGR has had
success with having a maintainer in an area where a single USD goes a
lot farther, and if we could find a potential maintainer in a position
like that, it might help us to spread our dollars.

Assuming a single silver sponsor -- or even 3, for example -- I agree
that the funds are limited enough that spending them on maintenance may
not be the best use of our money, in which case bunkers/code sprints
seem an ideal target, or other marketing efforts (for example, an
initial run of T-Shirts to give away at a conference or something like
that). 

> But seriously, my suggestion re the bunker thing was much more a sort
> of "easiest road to dog food" thing than necessarily a studied
> analysis of the problem/opportunities. If it's as easy as just dumping
> it into a bunker fund, that would be nice. Having to take decisions on
> who and how and where and when wrt the admin stuff.... although it may
> be very valuable...  does imply more work on our (PSC) part.

Sure. If the money pot is small enough, it's not worth it. If it gets
big enough, I think it might be a valuable position to have someone in
-- I've seen a lot of value from our ad-hoc efforts, and formalizing
these would be cool. 

> The final issue here is the PSC, and we seem to be in agreement. My
> only point there is that I think that if sponsors are going to be
> donating their money "to the project" (as you argue), essentially
> putting full faith in the PSC to take good care of it, then I think we
> need to take another look at our PSC and make sure it's even and fair
> enough that people who maybe aren't so familiar with the (stable,
> diligent) history of the PSC would be more (objectively) comfortable
> with it.

Absolutely. Establishing why our PSC is fair is likely something we
should do anyway :) But doing it for sponsorship is great. Extending
http://trac.openlayers.org/wiki/SteeringCommitteeMembers to explain some
of how we ensure the PSC represents the best interests of the project --
by adding, for example, that Schuyler is a long time participant in a
wide variety of Open Source communities, and has experience in working
to maintain and encourage these communities (rather than the current
"Schuyler is generally regarded as the loose cannon on deck.") would 
be a good step in this direction (as well as extending the PSC to 
include a more accurate representation of the project's participants).

> Oh and one more thing, re my suggestion for "guidelines" as to how to
> spend the money, which you seemed sort of vehemently opposed to. I
> didn't mean to suggest implying like rules for connecting donations to
> actual work... all I really meant was adding a line or two to the PSC
> page on the wiki or something like that with just some general rules
> about how we spend money. Really the only thing I had in mind was like
> a bullet point clearly explaining the decision process (vote,
> majority, unanimous, days to wait, etc) and maybe another bullet
> enumerating the different types of things that we'd consider spending
> money on.

Sure. This totally makes sense. I'd recommend we put it in the
sponsorship doc that I put in SVN, so we have a single place sponsors
can look. It already has the 'how we'll spend the money' bullets: we'd
just need to add the 'how we'll decide to spend the money' bullet points 
(which is partially/mostly just "Following the PSC rules for voting",
and partially something we need to craft to make these decisions.)  
If you don't feel like you can take this on (updating the doc), feel
free to just discuss prose on the list, and I'll put it in the doc when
we agree.

> Also keep in mind that one of the responsibilities is:
> 
> "To produce a detailed annual report of activities based on the
> sponsorship funding to be posted on the web site, and for distribution
> to sponsors. "
> 
> ...so there is another overhead that is added to the project. Would we
> nominate someone from the PSC to be in charge of this? Worth
> documenting.

It would clearly be a responsibility of the PSC to take care of this: if
we really couldn't get anyone to step up, the PSC would have to ensure
that it happened in some way (perhaps by sponsoring it, as unfortunate
as that would be. "Line Item: Spent money creating this report. 17% of
total expended funds." :))

> My whole issue with this whole thing is just that adding money into
> the mix I think always tends to complicate things. Even though I know
> we already *have* companies essentially giving us money in the form of
> allowing us time to work on openlayers... i just think that when it's
> actually real money people will act in different ways.

Absolutely. And as we get it, we'll see what we have to do to keep
people happy :)

> One more thing that was just brought up in discussion is how are we
> going to balance out the recognition given to paying sponsors versus
> companies that simply donate developer time? I'm thinking specifically
> of MetaCarta, OpenGeo, Camptocamp, etc. These are all companies making
> huge investments in the project and yet getting no overt (logo, etc)
> recognition. If I were one of them I would feel that somewhat unfair.

I think that we should have something very similar to GDAL's credits
page: 

  http://gdal.org/credits.html

Sponsors are listed first, and have logos and project/company
descriptions. Corporate supporters are listed second with a comapny
name, link, and what they've helped to do. (OpenGeo, Camptocamp, and
MetaCarta all come to mind as corporate sponsors, each with a list of
past and ongoing contributions, such as "Worked on original Vector
work", "Ongoing support through part time support of two developers",
"offer/maintain web hosting", etc.)  

Thanks for the comments/questions, and glad we could talk this over in
person as well; hopefully this helps to answer some of the questions
attached.

Regards,
-- 
Christopher Schmidt
MetaCarta